Health literacy is broadly defined as the cognitive skills, background knowledge, social abilities, and motivation to understand health information and to use it in promoting and maintaining good health. The relation between poor health literacy and adverse health outcomes is well documented. In this application, we focus on the aspect of health literacy that is related to the comprehension of written health texts. The abilit to process complex, written information is essential if individuals are to make informed decisions regarding their physical, emotional, and financial well-being. Reading involves complex interactions between reader-related factors (e.g., word-decoding skill, working-memory capacity) and text-related properties (e.g., syntactic complexity, coherence). Our aim in this application is to apply a psycholinguistic and cognitive perspective to the study of health literac. It specifically addresses two of the goals that are described in the program announcement: we analyze and measure the mix of abilities and skills required to be functionally 'health literate' and we examine the best methods of developing and disseminating effective information sources and materials for audiences with different levels of health literacy. Our proposed study extends research on health literacy in two important ways. First, we examine specific text manipulations to obtain information about which are best able to improve comprehension in individual readers, and second, how common tests of medical knowledge are related to other reader characteristics. To date, texts have been manipulated using readability formulas (reducing the length of sentences and the frequency of words). Our study is based on the assumption that 'one-size-fits-all' manipulations of readability do not take into account the strengths and weaknesses of different readers, and may, in fact, diminish comprehension for some readers, as recent research on individual differences in reading comprehension has suggested. We base our manipulations of text difficulty on psycholinguistic research that has investigated the factors that influence specific subcomponents of reading. Word frequency is only one of these factors. Important variables have been identified at the sentence level (e.g., level of syntactic embedding, use of syntactic structures to highlight the importance of a linguistic element in a sentence) and variables at the text level (e.g., the explicitness of anaphoric reference, the number of new entities that must be integrated into the reader's representation of a text). We investigate how these manipulations differentially affect readers with varying linguistic and cognitive profiles. Second, we examine how common tests of medical knowledge are related to other reader characteristics. This is important because knowledge and basic reading skills are often correlated. Readers who score low on background knowledge measures, tend to have poorer word-decoding skills, have lower scores on measures of working-memory capacity, have less reading experience, and poorer word knowledge than do readers who score high on these measures. This makes is difficult to determine which abilities are central to literacy and which are derivative. Our aim is to disentangle these effects, enabling a more focused approach to manipulations of text readability and health literacy. Our application has two phases. Phase one involves investigating (1) the basic linguistic and cognitive abilities of a large sample of readers, (2) the text properties that influence the difficlty of a text, and (3) how these factors interact. We apply a statistical technique called multilevel modeling to investigate how readers' unique abilities (e.g., word-decoding skill, WMC, reading experience) combine with text properties (e.g., word frequency, syntactic difficulty, anaphoric reference) to influence on-line reading processes how these influence subsequent memory for text information. The second phase of our study is designed to examine the comprehension performance of different groups of readers when specific characteristics of texts are manipulated. We have preliminary data from a sample of young adults suggesting that sentence length, the number of new concepts in a sentence, semantic overlap, and syntactic complexity exert a large influence on reading processes and subsequent comprehension. Importantly, these factors do not have uniform effects on reading. Rather, their influence varies as a function of working-memory capacity, decoding ability, and reading experience that are known to vary among individual readers.